Loading...
Paper Type
ERF
Abstract
Digital platforms increasingly reprocess user-generated content through preference filtering, aggregation, and AI-generated summaries. Although such algorithmic intermediation is widely viewed as a decision aid that enhances user experience, it may also influence how users interpret information presented to them. This study examines how presentation frames shape users’ decision confidence and investigates skepticism as an underlying mechanism. Using an online experiment (N = 200), we manipulate the framing of identical reviews as either aggregate summaries or preference-filtered content. Results show that preference-filtered presentation significantly reduces decision confidence relative to aggregate framing. Mediation analyses indicate that this effect operates through heightened skepticism, reflecting users’ questioning of information completeness and system intent. By introducing decision confidence as a process-oriented outcome and highlighting skepticism as a response to algorithmic intermediation, this study identifies a psychological cost of preference filtering and offers implications for research on algorithmic information environments and platform design in AI-mediated contexts.
Paper Number
1201
Recommended Citation
Zhong, Yunzhou and Deng, Bingjie, "When Personalization Backfires: Skepticism and Decision Confidence in Platform-Mediated Information" (2026). AMCIS 2026 Proceedings. 2.
https://aisel.aisnet.org/amcis2026/sig_hci/sig_hci/2
When Personalization Backfires: Skepticism and Decision Confidence in Platform-Mediated Information
Digital platforms increasingly reprocess user-generated content through preference filtering, aggregation, and AI-generated summaries. Although such algorithmic intermediation is widely viewed as a decision aid that enhances user experience, it may also influence how users interpret information presented to them. This study examines how presentation frames shape users’ decision confidence and investigates skepticism as an underlying mechanism. Using an online experiment (N = 200), we manipulate the framing of identical reviews as either aggregate summaries or preference-filtered content. Results show that preference-filtered presentation significantly reduces decision confidence relative to aggregate framing. Mediation analyses indicate that this effect operates through heightened skepticism, reflecting users’ questioning of information completeness and system intent. By introducing decision confidence as a process-oriented outcome and highlighting skepticism as a response to algorithmic intermediation, this study identifies a psychological cost of preference filtering and offers implications for research on algorithmic information environments and platform design in AI-mediated contexts.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
SIG HCI